The most commonly used anode materials for lithium-ion batteries are natural graphite and synthetic graphite (or artificial graphite) that can be intercalated with lithium and the resulting graphite intercalation compound (GIC) may be expressed as LixC6, where x is typically less than 1. The maximum amount of lithium that can be reversibly intercalated into the interstices between graphene planes of a perfect graphite crystal corresponds to x=1, defining a theoretical specific capacity of 372 mAh/g.
Graphite or carbon anodes can have a long cycle life due to the presence of a protective surface-electrolyte interface layer (SEI), which results from the reaction between lithium and the electrolyte (or between lithium and the anode surface/edge atoms or functional groups) during the first several charge-discharge cycles. The lithium in this reaction comes from some of the lithium ions originally intended for the charge transfer purpose. As the SEI is formed, the lithium ions become part of the inert SEI layer and become irreversible, i.e. they can no longer be the active element for charge transfer. Therefore, it is desirable to use a minimum amount of lithium for the formation of an effective SEI layer. In addition to SEI formation, the irreversible capacity loss Qir can also be attributed to graphite exfoliation caused by electrolyte/solvent co-intercalation and other side reactions.
In addition to carbon- or graphite-based anode materials, other inorganic materials that have been evaluated for potential anode applications include metal oxides, metal nitrides, metal sulfides, and the like, and a range of metals, metal alloys, and intermetallic compounds that can accommodate lithium atoms/ions or react with lithium. Among these materials, lithium alloys having a composition formula of LiaA (A is a metal such as Al, and “a” satisfies 0<a #5) are of great interest due to their high theoretical capacity, e.g., Li4Si (3,829 mAh/g), Li4.4Si (4,200 mAh/g), Li4.4Ge (1,623 mAh/g), Li4.4Sn (993 mAh/g), Li3Cd (715 mAh/g), Li3Sb (660 mAh/g), Li4.4Pb (569 mAh/g), LiZn (410 mAh/g), and Li3Bi (385 mAh/g). However, in the anodes composed of these materials, severe pulverization (fragmentation of the alloy particles) occurs during the charge and discharge cycles due to expansion and contraction of the anode active material induced by the insertion and extraction of the lithium ions in and out of the anode active material. The expansion and contraction, and the resulting pulverization of active material particles lead to loss of contacts between active particles and conductive additives and loss of contacts between the anode active material and its current collector. This degradation phenomenon is illustrated in FIG. 1. These adverse effects result in a significantly shortened charge-discharge cycle life.
To overcome the problems associated with such mechanical degradation, three technical approaches have been proposed:    (1) reducing the size of the active material particle, presumably for the purpose of reducing the strain energy that can be stored in a particle, which is a driving force for crack formation in the particle. However, a reduced particle size implies a higher surface area available for potentially reacting with the liquid electrolyte. Such a reaction is undesirable since it is a source of irreversible capacity loss.    (2) depositing the electrode active material in a thin film form directly onto a current collector, such as a copper foil. However, such a thin film structure with an extremely small thickness-direction dimension (typically much smaller than 500 nm, often necessarily thinner than 100 nm) implies that only a small amount of active material can be incorporated in an electrode (given the same electrode or current collector surface area), providing a low total lithium storage capacity and low lithium storage capacity per unit electrode surface area (even though the capacity per unit mass can be large). Such a thin film must have a thickness less than 100 nm to be more resistant to cycling-induced cracking, further diminishing the total lithium storage capacity and the lithium storage capacity per unit electrode surface area. Such a thin-film battery has very limited scope of application. A desirable and typical electrode thickness is from 100 μm to 200 μm. These thin-film electrodes (with a thickness of <500 nm or even <100 nm) fall short of the required thickness by three (3) orders of magnitude, not just by a factor of 3.    (3) using a composite composed of small electrode active particles protected by (dispersed in or encapsulated by) a less active or non-active matrix, e.g., carbon-coated Si particles, sol gel graphite-protected Si, metal oxide-coated Si or Sn, and monomer-coated Sn nano particles. Presumably, the protective matrix provides a cushioning effect for particle expansion or shrinkage, and prevents the electrolyte from contacting and reacting with the electrode active material. Examples of anode active particles are Si, Sn, and SnO2. Unfortunately, when an active material particle, such as Si particle, expands during the battery charge step, the protective coating is easily broken due to the mechanical weakness and/o brittleness of the protective coating materials. There has been no high-strength and high-toughness material available that is itself also lithium ion conductive.
It may be further noted that the coating or matrix materials used to protect active particles (such as Si and Sn) are carbon, sol gel graphite, metal oxide, monomer, ceramic, and lithium oxide. These protective materials are all very brittle, weak (of low strength), and/or non-conducting (e.g., ceramic or oxide coating). Ideally, the protective material should meet the following requirements: (a) The coating or matrix material should be of high strength and stiffness so that it can help to refrain the electrode active material particles, when lithiated, from expanding to an excessive extent. (b) The protective material should also have high fracture toughness or high resistance to crack formation to avoid disintegration during repeated cycling. (c) The protective material must be inert (inactive) with respect to the electrolyte, but be a good lithium ion conductor. (d) The protective material must not provide any significant amount of defect sites that irreversibly trap lithium ions. (e) The protective material must be lithium ion conductive. The prior art protective materials all fall short of these requirements. Hence, it was not surprising to observe that the resulting anode typically shows a reversible specific capacity much lower than expected. In many cases, the first-cycle efficiency is extremely low (mostly lower than 80% and some even lower than 60%). Furthermore, in most cases, the electrode was not capable of operating for a large number of cycles. Additionally, most of these electrodes are not high-rate capable, exhibiting unacceptably low capacity at a high discharge rate.
Due to these and other reasons, most of prior art composite electrodes have deficiencies in some ways, e.g., in most cases, less than satisfactory reversible capacity, poor cycling stability, high irreversible capacity, ineffectiveness in reducing the internal stress or strain during the lithium ion insertion and extraction steps, and other undesirable side effects.
Complex composite particles of particular interest are a mixture of separate Si and graphite particles dispersed in a carbon matrix; e.g. those prepared by Mao, et al. [“Carbon-coated Silicon Particle Powder as the Anode Material for Lithium Batteries and the Method of Making the Same,” US 2005/0136330 (Jun. 23, 2005)]. Also of interest are carbon matrix-containing complex nano Si (protected by oxide) and graphite particles dispersed therein, and carbon-coated Si particles distributed on a surface of graphite particles Again, these complex composite particles led to a low specific capacity or for up to a small number of cycles only. It appears that carbon by itself is relatively weak and brittle and the presence of micron-sized graphite particles does not improve the mechanical integrity of carbon since graphite particles are themselves relatively weak. Graphite was used in these cases presumably for the purpose of improving the electrical conductivity of the anode material. Furthermore, polymeric carbon, amorphous carbon, or pre-graphitic carbon may have too many lithium-trapping sites that irreversibly capture lithium during the first few cycles, resulting in excessive irreversibility.
In summary, the prior art has not demonstrated a composite material that has all or most of the properties desired for use as an anode material in a lithium-ion battery. Thus, there is an urgent and continuing need for a new anode for the lithium-ion battery that has a high cycle life, high reversible capacity, low irreversible capacity, small particle sizes (for high-rate capacity), and compatibility with commonly used electrolytes. There is also a need for a method of readily or easily producing such a material in large quantities.
In response to these needs, one of our earlier applications discloses a nano-scaled graphene platelet-based composite composition for use as a lithium ion battery anode [A. Zhamu and B. Z. Jang, “Nano Graphene Platelet-Based Composite Anode Compositions for Lithium Ion Batteries,” U.S. patent application Ser. No. 11/982,672 (Nov. 5, 2007); Now U.S. Pat. No. 7,745,047 (Jun. 29, 2010)]. This composition comprises: (a) micron- or nanometer-scaled particles or coating of an anode active material; and (b) a plurality of nano-scaled graphene platelets (NGPs), wherein a platelet comprises a graphene sheet or a stack of graphene sheets having a platelet thickness less than 100 nm and wherein the particles or coating are physically attached or chemically bonded to NGPs. Nano graphene platelets (NGPs) are individual graphene sheets (individual basal planes of carbon atoms isolated from a graphite crystal) or stacks of multiple graphene planes bonded together in the thickness direction. The NGPs have a thickness less than 100 nm and a length, width, or diameter that can be greater or less than 10 μm. The thickness is more preferably less than 10 nm and most preferably less than 1 nm.
Disclosed in another patent application of ours is a more specific composition, which is composed of a 3-D network of NGPs and/or other conductive filaments and select anode active material particles that are bonded to these NGPs or filaments through a conductive binder [Jinjun Shi, Aruna Zhamu and Bor Z. Jang, “Conductive Nanocomposite-based Electrodes for Lithium Batteries,” U.S. patent application Ser. No. 12/156,644 (Jun. 4, 2008)]. Yet another application, as schematically shown in FIGS. 2(A) and 2(B), provides a nano graphene-reinforced nanocomposite solid particle composition containing NGPs and electrode active material particles, which are both dispersed in a protective matrix (e.g. a carbon matrix) [Aruna Zhamu, Bor Z. Jang, and Jinjun Shi, “Nano Graphene Reinforced Nanocomposite for Lithium Battery Electrodes,” U.S. patent application Ser. No. 12/315,555 (Dec. 4, 2008)].
After our discovery of graphene providing an outstanding support for anode active materials, many subsequent studies by others have confirmed the effectiveness of this approach. For instance, Wang, et al. investigated self-assembled TiO2-graphene hybrid nanostructures for enhanced Li-ion insertion [D. Wang, et al. “Self-Assembled TiO2-Graphene Hybrid Nanostructures for Enhanced Li-Ion Insertion.” ACS Nano, 3 (2009) 907-914]. The results indicate that, as compared with the pure TiO2 phase, the specific capacity of the hybrid was more than doubled at high charge rates. The improved capacity at a high charge-discharge rate was attributed to increased electrode conductivity afforded by a percolated graphene network embedded into the metal oxide electrodes. However, all these earlier studies were focused solely on providing a network of electron-conducting paths for the anode active material particles and failed to address other critical issues, such as ease of anode material processing, electrode processability, electrode tap density (the ability to pack a dense mass into a given volume), and long-term cycling stability. For instance, the method of preparing self-assembled hybrid nanostructures is not amenable to mass production. The anode material particle-coated graphene sheets alone are not suitable for electrode fabrication (due to the difficulty in coating the materials onto a current collector), and the resulting electrodes are typically too low in the tap density. Additionally, paper-based composite structures are not compatible with current lithium-ion battery production equipment. These are all critically important issues that must be addressed in a real battery manufacturing environment.
Herein reported is a process for producing significantly improved anode composition that provides not only a robust 3-D network of electron-conducting paths and high conductivity, but also enables the anode materials to be readily made into electrodes with a high electrode tap density, a sufficiently large electrode thickness (typically 100-200 to ensure a sufficient amount of output current), a large weight percentage of anode active material (with respect to the amounts of non-active materials, such as a current collector and binder, in an electrode), and long-term cycling stability. Both the reversible capacity and the first-cycle efficiency are also significantly improved over those of state-of-the-art anode materials.